Real-time COVID-19 detection over chest x-ray images in edge computing.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computer Science

Background:

• The COVID-19 pandemic highlighted limitations in manual detection methods.
• Centralized deep learning models for COVID-19 detection face challenges in latency, privacy, and cost.

Purpose of the Study:

• To propose an efficient and accurate COVID-19 detection scheme using chest X-ray (CXR) images.
• To leverage edge computing and a lightweight Convolutional Neural Network (CNN) model to overcome limitations of centralized approaches.

Main Methods:

• A COVID-19 detection framework utilizing CXR image classification with the MobileNet model in an edge computing environment.
• Implementation of a lightweight CNN (MobileNet) for CXR image analysis.
• Utilizing a Deep Convolutional Generative Adversarial Network (DCGAN) for data augmentation to address small dataset sizes.

Main Results:

• The proposed scheme demonstrates efficient and accurate detection of COVID-19 from CXR images.
• Edge computing implementation alleviates computational burden on centralized data centers.
• MobileNet model shows effectiveness in classifying CXR images for COVID-19 detection.

Conclusions:

• The developed edge computing scheme offers a viable solution for rapid and accurate COVID-19 detection.
• Lightweight CNNs like MobileNet are suitable for efficient medical image analysis in decentralized settings.
• The approach improves upon traditional methods by addressing latency, privacy, and cost concerns.